[igt-dev] [PATCH i-g-t 2/7] meson: Don't redefine gettid if the C library provides it

[Petri Latvala]

On Tue, Apr 16, 2019 at 04:10:56PM -0400, Lyude wrote:
> From: Lyude Paul <lyude at redhat.com>
> 
> glibc 2.30+ will actually include a definition for gettid() that makes
> it so that users don't have to manually define a wrapper for it
> themselves with syscall(). We don't currently check for this, and as a
> result will end up redefining gettid() on the latest versions of glibc,
> causing the build to fail:
> 
> FAILED: lib/76b5a35@@igt-igt_kmod_c at sta/igt_kmod.c.o
> In file included from /usr/include/unistd.h:1170,
>                  from ../../mnt/vol/lib/igt_core.h:43,
>                  from ../../mnt/vol/lib/igt_kmod.c:28:
> /usr/include/bits/unistd_ext.h:34:28: error: macro "gettid" passed 1 arguments, but takes just 0
>    34 | extern __pid_t gettid (void) __THROW;
>       |                            ^
> In file included from ../../mnt/vol/lib/igt_kmod.c:27:
> ../../mnt/vol/lib/igt_aux.h:40: note: macro "gettid" defined here
>    40 | #define gettid() syscall(__NR_gettid)
>       |
> [36/771] Compiling C object 'lib/76b5a35@@igt-igt_kms_c at sta/igt_kms.c.o'.
> ninja: build stopped: subcommand failed.
> 
> So, fix this by by adding some meson checks to define HAVE_GETTID whenever the
> host defines its own gettid(), and avoid redefining gettid() when HAVE_GETTID is
> defined.
> 
> This fixes build intel-gpu-tools for me on Fedora Rawhide

Unless that's the package name on Fedora (and then, why), igt-gpu-tools :)


> 
> Signed-off-by: Lyude Paul <lyude at redhat.com>
> ---
>  lib/igt_aux.h | 3 +++
>  meson.build   | 3 +++
>  2 files changed, 6 insertions(+)
> 
> diff --git a/lib/igt_aux.h b/lib/igt_aux.h
> index 55392790..39fefcbf 100644
> --- a/lib/igt_aux.h
> +++ b/lib/igt_aux.h
> @@ -36,7 +36,10 @@
>  #include <i915/gem_submission.h>
>  
>  /* signal interrupt helpers */
> +#ifndef HAVE_GETTID
>  #define gettid() syscall(__NR_gettid)
> +#endif
> +
>  #define sigev_notify_thread_id _sigev_un._tid
>  
>  /* auxialiary igt helpers from igt_aux.c */
> diff --git a/meson.build b/meson.build
> index 557400a5..e73275dd 100644
> --- a/meson.build
> +++ b/meson.build
> @@ -237,6 +237,9 @@ if cc.has_header('cpuid.h')
>  	# FIXME: Do we need the example link test from configure.ac?
>  	config.set('HAVE_CPUID_H', 1)
>  endif
> +if cc.has_header_symbol('unistd.h', 'gettid', args : '-D_GNU_SOURCE')
> +	config.set('HAVE_GETTID', 1)
> +endif


All users of gettid() have #include unistd.h? Check.

Are there any other direct calls to syscall(__NR_gettid)? Or
syscall(SYS_gettid)? (is that the same?)

benchmarks/gem_syslatency.c:#define gettid() syscall(__NR_gettid)
lib/igt_core.c:	      	       pid_t tid = syscall(SYS_gettid);
tests/core_auth.c:	       return syscall(SYS_gettid) == tid;
tests/drm_import_export.c:     igt_debug("start %ld\n", syscall(SYS_gettid));
tests/i915/gem_close_race.c:#define gettid() syscall(__NR_gettid)



-- 
Petri Latvala